Current Limitations

Current Limitations

This page records verified implementation gaps and explicit non-claims. It is kept consistent with docs/PLATFORM-SUPPORT.json, project trackers, readiness documents, and the QEMU release-candidate contract.

Platform and hardware

• AArch64 QEMU provides the broadest complete OS-service path. QEMU validates behavior, not physical ARM performance, firmware behavior, or scaling.
• VMware Fusion 25.0.1 on Apple Silicon reaches /init through a generated ARM64 compatibility stage. VMware networking, persistent storage, multi-vCPU discovery, and later service gates are not integrated.
• The x86_64 image starts MADT-discovered processors, performs a controlled INT3 round trip and local-APIC timer interrupt, executes a ring-3 syscall, validates runtime-sized XSAVE and ACPI data, and exercises modern VirtIO block DMA/MSI-X plus network TX.
• Full AArch64 service parity on x86_64 remains open: complete userspace and thread services, receive networking and SSH, mounted filesystems, x86 NVMe operation, process-owned security services, AI Cell integration, and telemetry are absent.
• Physical Apple, Intel desktop, Xeon, SMMU/IOMMU, NVMe, NIC, NUMA, many-core, thermal, power, and performance evidence is not present.

Networking and SSH

• FreeBSD 15.1, native macOS, and Debian 13 OpenSSH clients pass bounded QEMU interoperability suites. This is not a production Internet deployment or an independent security audit.
• The SSH service deliberately supports four transports and two active channels per transport. Fleet-scale identity, audit, replay, and connection policy remains unresolved.
• SSH key exchange is classical curve25519-sha256 only. Hybrid post-quantum key exchange, OpenSSH interoperability and downgrade-policy review remain production security gates; the quiet local QEMU launcher only suppresses the OpenSSH 10 client notice for that development connection.
• DNS performs asynchronous external A-record resolution with timeout, retry, cache, and a QEMU-verified cache hit. DNSSEC, TCP fallback, complete AAAA application results, and deployment resolver policy remain absent.
• TCP implements retained segments, cumulative and partial ACK handling, RTT/RTO backoff, SACK, fast retransmit, zero-window handling, bounded reordering, keepalive, and FIN bookkeeping. Repeated-loss physical-network soak and congestion-control tuning remain unverified.
• Bounded IPv4/IPv6 fragment reassembly passes focused cases. Broad hostile fragment fuzzing remains incomplete.

Storage and persistence

• VirtIO block/network use interrupt-driven completions, event-index suppression, indirect descriptors, and bounded queued work. Emulated NVMe covers focused identify/write/flush/read and backing-byte checks.
• Production NVMe multiqueue, queue affinity, cancellation, direct final-buffer expert reads, physical durability, discard behavior, and throughput remain open.
• ModelFS supports signed registration, resumable staging, verification, immutable activation, scrub/quarantine, cleanup/reuse, and free-only trim under hosted and QEMU tests.
• Trusted-replica repair, production signing and key custody, physical multi-terabyte transfer, and model-v2 execution admission are not complete.
• ModelFS activation and MutableFS audit persistence are separate durability domains. A post-publication audit failure cannot roll back an already published active generation.

Administration and security

• xaios.control.v1 operations are bounded to 16 active keys, 16 revoked fingerprints, 64 audit/replay records, and 16 shell contexts. These are implementation limits, not fleet-scale targets.
• Role, capability, replay, rollback, host-key rotation, sensitive-path denial, and secret-redaction behavior pass QEMU/OpenSSH gates but have not received an independent production security review.
• Update signing uses a development trust root to validate transaction, fallback, and rollback behavior. Production key management is unresolved.

Inference engine and model support

• The kernel model-v1 path is a deterministic fixture. It does not execute a transformer and must not be described as real inference.
• Model-v2 parsing, streaming writing, architecture/backend registries, immutable readers, sessions, and scalar packed kernels are foundations only. Model-v2 packages are not yet executed end to end.
• No official tokenizer importer, real Qwen tensor importer, transformer plan, logits parity, or deterministic 32-token decode parity exists.
• Qwen 3.6 27B is blocked until the XAIOS platform completion gate. Kimi K3, DeepSeek V4 Flash 0731, and GLM 5.2 remain later roadmap targets.
• Kimi K3 KDA, Gated MLA, AttnRes, exact top-16 routing, shared experts, native MXFP4, text parity, and multimodal execution are not implemented.
• Scalar INT4/INT6 and experimental NEON/AVX2 packed kernels pass bounded correctness tests. Physical AVX2 validation, AVX-512/VNNI, AMX, SVE, tiled prefill/verification, persistent worker gangs, and bandwidth autotuning remain incomplete.
• No complete model-executing native macOS process, Metal backend, physical model-parity run, cluster data plane, or immutable performance artifact exists.

Evidence policy

Sparse files above 4 GiB or 100 GiB prove address width and bounded-memory behavior, not physical transfer throughput. High-vCPU QEMU gates prove dynamic metadata capacity, not server scalability. Performance claims require physical artifacts satisfying docs/BENCHMARK-CONTRACT.md.

See Open Decisions, Platform Support, and Model Support Roadmap.